Abstract

It is shown that unavoidable uncertainties arising from the experimental conditions in which systems are prepared for Einstein-Podolsky-Rosen experiments severely limit the possibilities for prediction. In the example originally proposed by EPR, time measurements are necessary for precise position predictions. If the preparation is designed to make the timing errors negligible, the parameters chosen for the preparation fix minimum uncertainties in the predictions leaving the observer no choice in the matter. In the case of correlated spin measurements, the preparation leaves the observer no control over the effect of the detector on the particles; the uncertainty of the previous condition of any one particle is so large that no (classical) prediction of the response of another detector is possible. Thus, there are as yet no proposed experiments which satisfy the conditions required by EPR—i.e., precise predictions which violate the uncertainty principle